Direct Effects of Model 3 - AMOS RETAINED MODEL OF THE MODERATING ROLE OF SELF-

S. AMOS RETAINED MODEL OF THE MODERATING ROLE OF SELF-

12. Direct Effects of Model 3

purposes of parsimony, however, paths are discussed in small, thematic groups in this section.

The relationships between the two indicators of body responsiveness, mind-body congruence and mind-body incongruence, and other consequences for women’s subjective experiences were modified first. Direct paths were added from body shame to mind-body congruence (standardized residual covariance = -3.34) and mind-body incongruence (residual covariance = 7.54); it is likely that higher levels of body shame would decrease one’s

motivation and/or ability to cue into internal bodily signals. This link is supported by the findings of Tylka and Hill (2004), who demonstrated that body shame predicted unique variance in poor awareness of hunger, satiety, and emotions. The confidence about bodily messages inherent in mind-body congruence is inconsistent with high levels of body-related anxiety, suggesting a negative effect of appearance anxiety on mind-body congruence (residual covariance = -5.54) and an exacerbating effect on mind-body incongruence (residual covariance = 8.02). Direct pathways were added from the body responsiveness variables to sense of control (residual covariance = 8.82 for mind-body congruence and residual covariance = -4.57 for mind-body incongruence). This aspect of dispositional flow taps into one’s feelings of control over his or her body; it is likely that sense of control is affected by one’s ability to sense and respond to their bodily feelings. Similarly, direct pathways were added from the body responsiveness variables to loss of self-consciousness; lack of mind-body congruence (residual covariance = 4.94) and mind-body incongruence (residual covariance = 1.95) may contribute to an increased reliance on others’ evaluations of one’s body and performance. Finally, the body responsiveness variables were allowed to

correlate with concentration due to the overlaps with the other dispositional flow variables and high standardized residual covariances (7.40 for mind-body congruence and -3.84 for mind-body incongruence). Fit statistics for this alternate model were χ2(17) = 98.49, p < .001, CFI = .96, RMSEA = .10, 90% CI [.08, .12], and SRMR = .06. Although this model fits the data significantly better than the original model, Δχ2(10) = 278.05, p < .001, it is still a poor global fit based on Kline’s (2016) criteria. In addition, several standardized covariance residuals were still greater than |2.00|.

The next set of modifications were made around appearance anxiety based on

findings in Grotewiel and Marszalek’s (2013) study. In this study, social appearance anxiety was found to be associated with loss of self-consciousness, sense of control, and

concentration as well or better than body surveillance. Drawing from work on the role of affect on evaluations of experiences (e.g., Klaaren, Hodges, & Wilson, 1994), it was assumed that affect (in this case, appearance anxiety) would influence whether or not a person

experiences flow. For example, a woman therapist who fears that her appearance is being evaluated by a client may have difficulty getting into flow because she appraises the situation as threatening. These effects are likely to carry over to the trait level. Thus, direct pathways were added from appearance anxiety to concentration (standardized residual covariance = - 3.31), control (standardized residual covariance = -3.89), and loss of self-consciousness (standardized residual covariance = -3.25). Fit statistics for this alternate model were χ2(14) = 59.95, p < .001, CFI = .98, RMSEA = .08, 90% CI [.06, .10], and SRMR = .04. This model again fit the data significantly better than the previous model, Δχ2(3) = 38.54, p < .001, but it still had poor local fit and mixed indicators of global fit.

The final set of modifications were made around physical safety anxiety. Like body responsiveness, physical safety anxiety has received relatively less empirical attention than other consequences for women’s subjective experience. Theoretically, it is possible that women who have experienced threats to their physical safety (i.e.., being raped, assaulted, attacked, robbed, or mugged) will have higher levels of physical safety anxiety as well as higher levels of shame. Thus, physical safety anxiety and body shame were allowed to correlate (standardized residual covariance = 2.91). Furthermore, high levels of physical safety anxiety are likely associated with high levels of anxiety across the board, including social appearance anxiety, so physical safety anxiety and social appearance anxiety were allowed to correlate (standardized residual covariance = 3.15). A path was added from physical safety anxiety to concentration due to its distracting effect (standardized residual covariance = 2.28). Finally, physical safety anxiety was allowed to correlate with the body responsiveness variables (for mind-body congruence, 2.70 and for mind-body incongruence, 2.64). It is likely that a reciprocal effect exists between physical safety anxiety and body responsiveness, such that high levels of physical safety anxiety (including vigilance) make listening and responding to one’s own body more difficult, whereas high levels of body responsiveness (e.g., feeling in touch with and control over one’s own body) may decrease physical safety anxiety. Fit statistics for this alternate model were χ2(9) = 16.17, p > .05, CFI = 1.0, RMSEA = .04, 90% CI [.00, .07], and SRMR = .01. This model again fit the data significantly better than the previous model, Δχ2(5) = 43.78, p < .001. Global model fit was improved overall. Local fit was also improved, with no standardized covariance residuals

greater than |2.00|. This final model, with a total addition of 18 paths and correlations, was retained.

Summary of retained Model 1. Departures from the original Model 1 include the addition of paths from body shame to (1) mind-body congruence and (2) mind-body incongruence; appearance anxiety to (3) mind-body congruence and (4) mind-body incongruence, as well as (5) concentration, (6) control, and (7) loss of self-consciousness; mind-body congruence to (8) control and (9) loss of self-consciousness; mind-body

incongruence to (10) control and (11) loss of self-consciousness; and physical safety anxiety to (12) concentration. Correlations were added between (1) mind-body congruence and concentration; (2) mind-body incongruence and concentration; (3) physical safety anxiety and body shame; (4) physical safety anxiety and social appearance anxiety; (5) physical safety anxiety and mind-body congruence; and (6) physical safety anxiety and mind-body incongruence.

Combined, all predictors in this final model accounted for 48% of the variance in disordered eating, 43% of the variance in depression symptoms, and 13% of the variance in sexual functioning. Variance estimates and squared multiple correlations (R2) for all variables are summarized in Table 5. Using Cohen’s (1988) guidelines for interpreting effect size (i.e., .01 = small, .09 = medium, .25 = large), the combined predictors had a large effect on

disordered eating and depression symptoms, and a medium effect on sexual functioning. Most direct effects were significant with small or medium effect sizes (see Table 6).

Seventeen hypothesized paths were not supported by the model. These paths were from body surveillance to (1) control and (2) physical safety anxiety; body shame to (3) mind-body

Table 5

Variance Estimates of the Final Path Model 1

Variable Unstandardized coefficient (SE) R2

OBC-Surv 1.51** (0.10) error 1 (OBC-Shame) 1.35** (0.09) .22 error 2 (SAAS) 184.56** (11.68) .19 error 3 (CONC) 8.43** (0.53) .09 error 4 (CONT) 6.26** (0.40) .28 error 5 (LOSS) 7.83** (0.50) .35 error 6 (BR-Con) 1.48** (0.09) .14 error 7 (BR-Incon) 1.58** (0.10) .21 error 8 (PSA) 70.83** (4.48) .01 disturbance 1 (EAT-26) 236.20** (14.95) .48 disturbance 2 (CES-D) 23.00** (1.45) .43 disturbance 3 (FSFI) 66.10** (4.18) .13

Note: OBC-Surv = Objectified Body Consciousness Scale, Body Surveillance subscale; OBC-Shame =

Objectified Body Consciousness Scale, Body Shame subscale; SAAS = Social Appearance Anxiety Scale; CONC = Dispositional Flow Scale-2 Long Form, Concentration subscale; CONT = Dispositional Flow Scale-2 Long Form, Control subscale; LOSS = Dispositional Flow Scale-2 Long Form, Loss of Self-Consciousness subscale; BR-Con = Daubenmier’s body responsiveness scale, mind-body congruence subscale; BR-Incon = Daubenmier’s body responsiveness scale, mind-body incongruence subscale; PSA = Physical safety anxiety scale; EAT-26 = Eating Attitudes Test-26; CES-D = Center for Epidemiologic Studies Depression Scale Short Form; FSFI = Female Sexual Function Index.

* = p < .05, ** = p < .01.

Table 6

Direct Path Coefficients of the Final Path Model 1

Path Unstandardized coefficient (SE) Standardized coefficient

OBC-Surv  OBC-Shame 0.47** (0.04) .44 OBC-Surv  SAAS 5.37** (0.52) .44 OBC-Surv  CONC -0.29* (0.12) -.12 OBC-Surv  CONT 0.01 (0.11) .00 OBC-Surv  LOSS -1.03** (0.13) -.36 OBC-Surv  BR-Con -0.12* (0.06) -.11 OBC-Surv  BR-Incon -0.12* (0.06) -.10 OBC-Surv  PSA 0.60 (0.32) .09 OBC-Shame  BR-Con -0.04 (0.06) -.04 OBC-Shame  BR-Incon 0.24** (0.07) .23 OCB-Shame  EAT-26 7.69** (0.09) .12 OCB-Shame  CES-D 0.06 (0.27) .01 OCB-Shame  FSFI -0.53 (0.40) -.08 SAAS  CONC -0.04** (0.01) -.21 SAAS  CONT -0.05** (0.01) -.27 SAAS  LOSS -0.05** (0.01) -.23 SAAS  BR-Con -0.03** (0.01) -.29 SAAS  BR-Incon 0.03** (0.01) .30 SAAS  EAT-26 0.17* (0.09) .12 SAAS  CES-D 0.12** (0.03) .28 SAAS  FSFI -0.05 (0.04) -.09 CONC  EAT-26 .57 (0.34) .08 CONC  CES-D -.24* (0.10) -.11 CONC  FSFI .05 (0.19) .02 CONT  EAT-26 0.36* (0.36) -.05 CONT  CES-D -0.45** (0.13) -.21 CONT  FSFI -0.38 (0.20) .12 LOSS  EAT-26 0.80** (0.27) .13 LOSS  CES-D 0.08 (0.08) .05 LOSS  FSFI -0.08 (0.15) -.03 BR-Con  CONT 0.77** (0.11) .34 BR-Con  LOSS 0.47** (0.12) .18 BR-Con  EAT-26 0.78 (0.68) .05 BR-Con  CES-D -0.18 (0.21) -.04 BR-Con  FSFI 1.39** (0.36) .21 BR-Incon  CONT -0.14 (0.10) -.07 BR-Incon  LOSS 0.05 (0.11) .02 118

Table 6--Continued

Path Unstandardized coefficient (SE) Standardized coefficient

BR-Incon  EAT-26 3.29** (0.59) .22 BR-Incon  CES-D 1.13** (0.19) .25 BR-Incon  FSFI 0.26 (0.30) .04 PSA  CONC 0.05** (0.01) .13 PSA  EAT-26 0.37** (0.09) .15 PSA  CES-D 0.10** (0.03) .14 PSA  FSFI -0.03 (0.05) -.02

Note: OBC-Surv = Objectified Body Consciousness Scale, Body Surveillance subscale; OBC-Shame =

* = p < .05, ** = p < .01

congruence, (4) depression symptoms, and (5) sexual functioning; appearance anxiety to (6) sexual functioning; control to (7) eating disorder symptoms; concentration to (8) eating disorder symptoms and (9) sexual functioning; loss of self-consciousness to (10) depression symptoms and (11) sexual functioning; mind-body congruence to (12) disordered eating and (13) depressive symptoms; mind-body incongruence to (14) control, (15) loss of self-

consciousness, and (16) sexual functioning; and physical safety anxiety to (17) sexual functioning (see Figure 5).

Mediation effects in retained model 1. In order to evaluate the mediation effects proposed in Hypotheses 1 through 3, the direct paths comprising these relationships were examined first. For cases in which all implicated direct effects were significant, Preacher and Leonardelli’s (2016) interactive calculation tool for the Aroian version of the Sobel (1982) test equation was used to examine significance of indirect effects. The Aroian version of the equation was chosen because it incorporates the standard error of the implicated regression coefficients (Preacher & Leonardelli, 2016). This is important because internal,

psychological mediators are likely to be measured with error; omitting the error terms often results in an underestimate of the effect of the mediator and an overestimate of the

independent variable (often an exogenous variable), ultimately resulting in successful mediation being falsely rejected (Baron & Kenny, 1986).

Hypothesis 1, that concentration would mediate relationships between body surveillance and health consequences, was not supported. Body surveillance had a small direct effect on concentration, b = -0.29, SE = 0.12, β = -.12, p < .05, and concentration had a small direct effect on depression symptoms, b = -0.24, SE = 0.10, β = -.11, p < .05. The

121

Figure 5. Retained model of the mediating role of three dimensions of flow (i.e., concentration, loss of self-consciousness, sense of control) in objectification theory. Solid lines indicate significant pathways (p < .05). Dotted lines indicate pathways that are not significant (p > .05). Bolded solid and dashed lines indicate pathways added to the proposed model in the modification process.

indirect effect of body surveillance on depression symptoms mediated by concentration was not significant, however, b = .07, SE = .04, β = .01, z = 1.64, p > .05. Furthermore, the practical significance of this indirect effect was very small. Concentration did not have a significant direct effect on disordered eating or sexual functioning. Hypothesis 2, that sense of control would mediate relationships between body surveillance and the mental health consequences, was not supported because body surveillance did not have a significant direct effect on sense of control. Hypothesis 3, that loss of self-consciousness would mediate relationships between body surveillance and the mental health outcomes, was partially supported. Body surveillance had a moderate direct effect on loss of self-consciousness, b = - 1.03, SE = 0.13, β = -.36, p < .01, and loss of self-consciousness had a small direct effect on disordered eating, b = 0.80, SE = 0.27, β = .13, p < .05. The indirect effect of body

surveillance on depression symptoms mediated by concentration was statistically significant but practically very small, b = -0.82, SE = .29, β = -.05, z = -2.77, p < .01. Loss of self- consciousness did not have a significant direct effect on depression symptoms or sexual functioning.

Model 2: The moderating role of dispositional mindfulness. Hypotheses 5-10 dealt with the moderating role of dispositional mindfulness within the objectification theory

framework. We initially hypothesized that dispositional mindfulness would moderate the links from body surveillance to concentration, control, loss of self-consciousness, and body responsiveness. We also hypothesized that dispositional mindfulness would moderate the mediating relationships between these variables (i.e., concentration, control, loss of self- consciousness, and body responsiveness) and the body surveillance-mental health outcome

(i.e., disordered eating, depression symptoms, and sexual functioning) links. Before testing, we amended the original hypothesized Model 2 (Figure 3) based on the modifications we made to Model 1. See Figure 6 for the amended hypothesized Model 2.

Model fit statistics were examined to verify the adequacy of Model 2, which was theoretically identified. Fit statistics for this model were poor: χ2(21) = 178.70, p < .001, CFI = .94, RMSEA = .12, 90% CI [.11, .15], and SRMR = .08. Several standardized residual covariances were greater than |2.00|, evidencing poor local fit. Together, these results suggested that the model should be modified to incorporate direct pathways or covariances with large standardized residual covariances in cases in which a relationship would be theoretically supported. Examination of standardized residual covariances suggested that there should be links from dispositional mindfulness to body shame (standardized residual covariance = -5.66) and appearance anxiety (standardized residual covariance = -7.69). These relationships are theoretically supported. Previous studies have demonstrated links between mindfulness and other body-related cognitive processes, including body comparison and body dissatisfaction (Dekeyser et al., 2008; Dijkstra & Barelds, 2011). Furthermore, mindfulness has been shown throughout multiple studies to negatively predict anxiety (see Brown et al., 2013). We also added pathways to check for moderation by dispositional mindfulness of the mediating effects of body shame (standardized residual covariance = - 1.34) and appearance anxiety (standardized residual covariance = -2.64) of body surveillance on the mental health outcomes. Finally, we also added paths from dispositional mindfulness to depression symptoms and sexual functioning. These pairs had large standardized residual covariances (-6.39 for mindfulness-depression and 3.59 for mindfulness-sexual functioning).

124

Conceptually, mindfulness has repeatedly been linked to depression (see Brown et al., 2013), and it is not surprising that concentration, sense of control, loss of self-consciousness, and body responsiveness may not fully mediate the relationship between these two variables. Likewise, mindfulness training has recently been explored as a treatment for women experiencing sexual dysfunction, with promising results (Brotto, Bassoon, & Luria, 2008; Brotto, Heiman, et al., 2008; Brotto et al., 2012; Silverstein et al., 2011); it may affect female functioning directly. Fit of this model was much better than the previous model, χ2_{(15) =} 27.59, p = .02, CFI = 1.0, RMSEA = .04, 90% CI [.02, .07], SRMR = .01, Δχ2(6) = 151.11, p < .001, with all standardized residual covariances less than |2.00|. However, a significant chi square value indicated that there was a significant difference between expected and observed covariance matrixes for the model, prompting further modification.

Modification indexes were checked for further recommendations to improve fit. The addition of a pathway from body surveillance to sexual functioning was suggested

(modification index = 4.65, parameter change = .61). When controlling for the moderating and mediating effects of dispositional mindfulness in the model, body surveillance had a direct effect on sexual functioning. In other words, body surveillance directly affects sexual functioning when all other variables in the model, including dispositional mindfulness and the interaction of dispositional mindfulness and body surveillance, are held constant. Fit of this model was much better than the previous model, with good global and local fit indices: χ2(14) = 20.72, p = .11, CFI = 1.0, RMSEA = .03, 90% CI [.00, .06], SRMR = .01, Δχ2(1) = 6.87, p < .01, all standardized residual covariances less than |2.00|. This final model was retained.

Summary of Retained Model 2. From the original hypothesized model, the modified Model 2 that was retained included the same changes made to Model 1 (i.e., the same addition of 18 paths and correlations). To the amended Model 2 that was initially tested, an additional seven paths were added: dispositional mindfulness to (1) body shame and (2) appearance anxiety; dispositional mindfulness to the mediating path from body surveillance to disordered eating, depression symptoms, and sexual functioning through (3) body shame and (4) appearance anxiety; dispositional mindfulness to (5) depression symptoms and (6) sexual functioning; and body surveillance to (7) sexual functioning.

Combined, all predictors in this final model accounted for 48% of the variance in disordered eating, 47% of the variance in depression symptoms, and 16% of the variance in sexual functioning. Variance estimates and squared multiple correlations (R2) for all variables are summarized in Table 7. The combined predictors had a large effect on disordered eating and depression symptoms and a medium effect on sexual functioning. With the interaction of body surveillance and dispositional mindfulness in the model, most direct effects were significant with small or medium effect sizes (see Table 8). Thirty hypothesized paths were not supported by the model. These paths were from body surveillance to (1) concentration, (2) control, (3) mind-body congruence, and (4) physical safety anxiety; the interaction of body surveillance and dispositional mindfulness to (5) body shame, (6) concentration, (7) control, (8) loss of self-consciousness, (9) mind-body congruence, and (10) mind-body incongruence; dispositional mindfulness to (11) mind-body incongruence; body shame to (12) mind-body congruence, (13) depression symptoms, and (14) sexual functioning; appearance anxiety to (15) concentration, (16) mind-body congruence, and (17) sexual

Table 7

Variance Estimates of the Final Path Model 2

Variable Unstandardized coefficient (SE) R2

OBC-Surv 1.51** (0.10) FMI-SF 58.6** (3.71) OBC-Surv x FMI-SF 96.45** (6.11) error 1 (OBC-Shame) 1.22** (0.08) .29 error 2 (SAAS) 152.00** (9.62) .33 error 3 (CONC) 6.68** (0.42) .28 error 4 (CONT) 5.22** (0.33) .40 error 5 (LOSS) 7.08** (0.45) .41 error 6 (BR-Con) 1.20** (0.08) .30 error 7 (BR-Incon) 1.58** (0.10) .21 error 8 (PSA) 70.83** (4.48) .01 disturbance 1 (EAT-26) 236.20** (14.95) .48 disturbance 2 (CES-D) 21.61** (1.37) .47 disturbance 3 (FSFI) 63.56** (4.02) .16

Note: OBC-Surv = Objectified Body Consciousness Scale, Body Surveillance subscale; FMI-SF = Freiburg

Mindfulness Inventory-Short Form; OBC-Shame = Objectified Body Consciousness Scale, Body Shame subscale; SAAS = Social Appearance Anxiety Scale; CONC = Dispositional Flow Scale-2 Long Form, Concentration subscale; CONT = Dispositional Flow Scale-2 Long Form, Control subscale; LOSS =

Dispositional Flow Scale-2 Long Form, Loss of Self-Consciousness subscale; BR-Con = Daubenmier’s body responsiveness scale, mind-body congruence subscale; BR-Incon = Daubenmier’s body responsiveness scale, mind-body incongruence subscale; PSA = Physical safety anxiety scale; EAT-26 = Eating Attitudes Test-26; CES-D = Center for Epidemiologic Studies Depression Scale Short Form; FSFI = Female Sexual Function Index.

* = p < .05, ** = p < .01.

Table 8

Direct Path Coefficients of the Final Path Model 2 Path

Unstandardized coefficient

(SE) Standardized coefficient

OBC-Surv  OBC-Shame 0.39** (0.05) .36 OBC-Surv  SAAS 4.04** (0.51) .33 OBC-Surv  CONC -0.14 (0.11) -.05 OBC-Surv  CONT 0.09 (0.10) .04 OBC-Surv  LOSS -0.97** (0.13) -.34 OBC-Surv  BR-Con -0.08 (0.05) -.07 OBC-Surv  BR-Incon -0.13* (0.06) -.12 OBC-Surv  PSA 0.60 (0.32) .09 OBC-Surv  FSFI 0.96** (0.35) .14 FMI-SF  OBC-Shame -0.05** (0.01) -.28 FMI-SF  SAAS -0.74** (0.08) -.37 FMI-SF  CONC 0.20** (0.02) .50 FMI-SF  CONT 0.17** (0.02) .44 FMI-SF  LOSS 0.14** (0.02) .32 FMI-SF  BR-Con 0.08** (0.01) .46 FMI-SF  BR-Incon -0.01 (0.01) -.06 FMI-SF  CES-D -0.25** (0.05) -.29 FMI-SF  FSFI 0.24** (0.07) .21

OBC-Surv x FMI-SF  OBC-Shame -0.01 (0.01) -.05

OBC-Surv x FMI-SF  SAAS -0.15* (0.06) -.10

OBC-Surv x FMI-SF  CONC 0.01 (0.01) .00

OBC-Surv x FMI-SF  CONT 0.00 (0.01) .00

OBC-Surv x FMI-SF  LOSS 0.00 (0.02) .03

OBC-Surv x FMI-SF  BR-Con 0.01 (0.01) .07

OBC-Surv x FMI-SF  BR-Incon 0.00 (0.01) .03

OBC-Shame  BR-Con 0.00 (0.06) .00 OBC-Shame  BR-Incon 0.24** (0.07) .23 OCB-Shame  EAT-26 7.69** (0.91) .48 OCB-Shame  CES-D -0.04 (0.25) -.01 OCB-Shame  FSFI -0.65 (0.40) -.10 SAAS  CONC 0.00 (0.01) .01 SAAS  CONT -0.03** (0.01) -.14 SAAS  LOSS -0.03** (0.01) -.14 SAAS  BR-Con -0.01 (0.01) -.10 SAAS  BR-Incon 0.03** (0.01) .29 SAAS  EAT-26 0.17* (0.09) .12 128

Table 8--Continued

Path Unstandardized Coefficient

(SE) Standardized coefficient SAAS  CES-D 0.10** (0.03) .24 SAAS  FSFI -0.05 (0.04) -.08 CONC  EAT-26 0.57 (0.34) .08 CONC  CES-D -0.15 (0.11) -.07 CONC  FSFI -0.01 (0.19) .00 CONT  EAT-26 -0.38 (0.36) -.05 CONT  CES-D -0.30* (0.13) -.14 CONT  FSFI 0.15 (0.20) .05 LOSS  EAT-26 .08** (0.27) .13 LOSS  CES-D 0.17* (0.08) .09 LOSS  FSFI -0.03 (0.15) -.01 BR-Con  CONT 0.37** (0.11) .17 BR-Con  LOSS 0.13 (0.13) .05 BR-Con  EAT-26 0.78 (0.68) .05 BR-Con  CES-D 0.18 (0.21) .04 BR-Con  FSFI 1.09** (0.37) .17 BR-Incon  CONT -.09 (0.09) -.04 BR-Incon  LOSS .09 (0.11) .04 BR-Incon  EAT-26 3.29** (0.59) .22 BR-Incon  CES-D 1.11** (0.18) .25 BR-Incon  FSFI 0.35 (0.30) .06 PSA  CONC 0.04** (0.01) .11 PSA  EAT-26 0.37** (0.09) .15 PSA  CES-D 0.09** (0.03) .12 PSA  FSFI -0.02 (0.05) -.02

Note: OBC-Surv = Objectified Body Consciousness Scale, Body Surveillance subscale; FMI-SF = Freiburg

* = p < .05, ** = p < .01.

functioning; concentration to (18) eating disorder symptoms, (19) depression, and (20) sexual functioning; control to (21) eating disorder symptoms and (22) sexual functioning; loss of self-consciousness to (23) sexual functioning; mind-body congruence to (24) loss of self- consciousness, (25) eating disorder symptoms, and (26) sexual functioning; mind-body incongruence to (27) control, (28) loss of self-consciousness, and (29) sexual functioning; and physical safety anxiety to (30) sexual functioning (see Figure 7).

Moderation effects. In order to evaluate the moderation effect proposed in

Hypothesis 5 and the mediated moderation effects proposed in Hypotheses 6 through 10, the direct effects of the interaction of body surveillance and dispositional mindfulness on

concentration, control, loss of self-consciousness, mind-body congruence, and mind-body incongruence were first examined. Counter to these hypotheses, none of these proposed relationships achieved statistical significance. Dispositional mindfulness did not appear to moderate the effect of body surveillance on these variables or to moderate the mediating effect of these variables on the relationship between body surveillance and the mental health outcomes.

As part of the model modification process, moderation by dispositional mindfulness of the mediation of the effect of body surveillance on the mental health outcomes by body shame and appearance anxiety was also tested. The moderating effect of dispositional mindfulness on the relationship between body surveillance and body shame was not significant. Analysis revealed a significant but small moderating effect on the relationship

In document The roles of dispositional flow, dispositional mindfulness, and self-compassion in the Objectification Theory Framework (Page 124-200)